Fatigue crack growth prediction method for offshore platform based on digital twin

To accurately predict the crack growth trend of marine structures, this paper proposes a fatigue crack growth prediction method for offshore platforms based on digital twin. First, a digital twin model for offshore platforms is established, and the key technical procedures required for each part of the model are given. Subsequently, to implement consistency maintenance between the virtual model and the physical entity during the digital twin usage, a finite element surrogate model approach based on Gaussian process is performed, which integrates with the crack growth consistency maintenance strategy using dynamic Bayesian network. Finally, according to service condition characteristics of the offshore platform, a crack growth experiment under mixed-mode multi-step loading is designed to verify the effectiveness of the proposed method. The results show that the method not only reduces the influence of uncertain factors on crack growth prediction under complex loads, but also achieves accurate of crack growth through dynamic tracking.

Automated summary

This paper proposes a fatigue crack growth prediction method for offshore platforms based on digital twin, which establishes a digital twin model and utilizes a finite element surrogate model approach to ensure consistency between virtual and physical models. The method is validated through a crack growth experiment under mixed-mode multi-step loading, demonstrating reduced influence of uncertain factors and accurate crack growth prediction through dynamic tracking.